# Efficient data transfer to co-processor (GPGPU using OpenCL)

**URL:** <https://www.ruby-forum.com/t/efficient-data-transfer-to-co-processor-gpgpu-using-opencl/233578>\
**Category:** GNU Radio\
**Created:** [December 25, 2014, 5:08pm UTC](https://www.ruby-forum.com/t/efficient-data-transfer-to-co-processor-gpgpu-using-opencl/233578 "2014-12-25T17:08:19Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![addis\_a](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/a/c4cdca/32.png) [@addis\_a](https://www.ruby-forum.com/u/addis_a)\
**Post date:** [December 25, 2014, 5:08pm UTC](https://www.ruby-forum.com/t/efficient-data-transfer-to-co-processor-gpgpu-using-opencl/233578/1 "2014-12-25T17:08:19Z")

</div>

Hello,

I am experimenting with OpenCL and need to create a GNU Radio processing  
block where all the computations will happen inside a discrete GPU. More  
precisely, I am writing an OpenCL implementation of the polyphase  
channelizer.

The block needs to process 1.6 Gbps (50Msps@32bit). If I do everything  
in one sychronous block, it means that every time that the work()  
function is called, I have to transfer the samples from host-\>GPU,  
compute and transfer back from GPU-\>host. Indeed, this is not optimal as  
the PCI-E bus can have achieve throughput but also have high latency.  
The optimal situation would be feeding and consuming from the GPU in an  
asynchronous fashion (using DMA?) so that the GPU doesn’t stop  
processing during transfers.

I know there are some ongoing efforts to better handle co-processors in  
GNU Radio ([http://gnuradio.org/redmine/projects/gnuradio/wiki/Keystone2](http://gnuradio.org/redmine/projects/gnuradio/wiki/Keystone2),  
[https://gnuradio.org/redmine/projects/gnuradio/wiki/GRCon13Coprocessor](https://gnuradio.org/redmine/projects/gnuradio/wiki/GRCon13Coprocessor)).  
Unfortunately, this is work is barely usable now and requires  
modification of the GNU Radio runtime.  
There is also the GREX project which has all the features I need  
(extended buffer API/pinned memory); but the project is discontinued.

There are numerous papers about implementing a polyphase channelizer on  
a gpu, and some of them claim that they did use GNU Radio but to my  
knowledge none of them provide a reference implementation or explained  
how they handled the buffer management inside GNU Radio.

As a workaround, I want to create a hier\_block2 with one “pinned memory”  
DMA sink feeding my GPU and one DMA source. Do you think this option is  
valid? What would be my other options about this?

Any opinion on the subject is welcome. Thank you very much!